Detection of Adulterated Botanicals

The easiest method for detecting adulteration of a botanical material is to compare the sample in question with reference botanicals (either and authentic or historic materials). When considering the chemical composition, such a comparison can be based on chromatographic or un-fractionated spectral profiles. A sample is considered adulterated when the sample profile is statistically different from the reference profile. An adulterant may be a foreign compound (a non-endogenous addition), a spike (an endogenous addition), or a substitution (another botanical or complex organic mixture). Profiles can be statistically examined using analysis of variance (ANOVA) or principal component analysis of the reference botanicals (PCA). Comparisons may be targeted (the adulterant is known) or un-targeted (adulterant(s) not known). In general, targeted comparisons have the advantage of being more specific with better statistical limits while untargeted analyses require more breadth. Thus, targeted analyses will generally use a single extraction, separation, and detection method. The comparison may be based on a specific portion of the chromatogram or spectrum. Truly untargeted comparisons will have to employ multiple methods with different extraction solvents, separations, and detection techniques. For targeted analyses, samples may be prepared with known levels of adulteration to characterize the sensitivity of the analysis, i.e. what concentration of the adulterant is necessary to produce a deviation from the reference profile. We will show that Panax quinquefolius adulterated with 5% Panax ginseng can be detected. Un-targeted comparisons will have to be treated as multiple targeted analyses.